What is used test equipment actually worth?
Calibration is priced by the work an instrument requires — bench time, test points, reference standards, paperwork. It is not priced as a fraction of what the instrument is worth.
That sounds too obvious to matter. Follow it across a price ladder and it stops being obvious. A calibration bill that barely registers against a high-end oscilloscope can eat an eighth of what a bench multimeter sells for.
Published advice on what used test equipment is worth tends to start and end with the calibration certificate. What it skips is the arithmetic: what that certificate costs relative to what the instrument actually fetches.
What used test equipment is worth, across six instruments
Custom-Cal, an ISO 17025-accredited lab, is unusual in publishing both sides of this: its used inventory carries an asking price and a per-unit calibration figure on the same listing. That permits a comparison across the range using one lab’s numbers.
| Instrument | Used price | Calibration as % of price |
|---|---|---|
| Keithley 197A microvolt DMM | $475 | 12.6% |
| Agilent 34401A 6½-digit DMM | $700 | 10.7% |
| Agilent 54845A 1.5 GHz Infiniium scope | $3,450 | 5.1% |
| Agilent 8595E 6.5 GHz spectrum analyzer | $4,200 | 8.9% |
| Tektronix TDS7104 1 GHz DPO | $5,000 | 4.5% |
| Agilent DSOX92004A 20 GHz Infiniium | $78,500 | 1.2% |
Sources: used multimeter, spectrum analyzer and oscilloscope listings, retrieved August 2026.
The spectrum analyzer breaks the pattern, and it’s worth saying why rather than quietly reordering the table. Spectrum analyzer calibration is dearer than oscilloscope calibration at any given price point — more test points, RF reference standards, longer on the bench. The same lab’s retail card puts spectrum analyzers at $325 to $1,490 against $145 to $1,180 for scopes.
So the ladder measures price and instrument class at once. Within a single class the relationship is clean: the two scopes run 5.1% at $3,450 and 1.2% at $78,500. Across classes, expect the sort of jump the 8595E shows.
Two further caveats belong with that table. All six figures come from one lab, and the listings don’t state whether the calibration figure is an add-on charge or the itemized cost of a calibration already included in the sale price.
The second caveat is worth testing rather than waving at, because it runs the wrong way for the argument. These per-unit figures are roughly half the same lab’s published retail card, where multimeter calibration runs $110 to $825.
So price the $475 Keithley against the $110 retail floor and the ratio climbs to 23%. Price the 20 GHz Infiniium against the $1,180 ceiling and it reaches 1.5%. The spread gets wider under retail pricing, not narrower.
Worth noting what those prices don’t include. The rate card states that pricing “does not in most cases include measurement performance data” — the as-found and as-left numbers a quality auditor would want. A certificate still carries real information: calibration date, traceability, tolerance status. But at this lab, the full data package is a separate line item, so “has a cal cert” and “has as-found data” are not the same claim.
What the certificate actually buys
Here the usual advice breaks down, and not in the direction sellers expect.
The large refurbishers recalibrate on intake regardless of what arrived. Keysight’s Premium Used refurbishment puts calibration at step six of ten. Tektronix Encore units are “tested, calibrated and reconditioned” to factory specification. TRS-RenTelco routes all used stock through its own accredited lab.
That doesn’t describe every buyer. A lab buying peer-to-peer, or anyone who needs an instrument in service the week it lands, may genuinely prefer to pay for a current certificate rather than budget for one. But the volume buyers — the ones most likely to quote on a mixed lot of surplus — are paying for a calibration either way.
So for that channel, a fresh certificate does not save the buyer a cost. It buys information instead.
An instrument that just passed calibration measured within spec on the day it was tested. That narrows the buyer’s risk without eliminating it — the unit can still have a dying fan, a worn encoder, or an intermittent nobody caught in an afternoon on the bench. Narrowing is still worth something, because measured performance is the part a buyer cannot assess from photographs. On a $78,500 scope, spending 1.2% to narrow that doubt is trivially worth it. On a $475 meter, spending an eighth of the sale value to send a signal the buyer will overwrite anyway is harder to justify.
That is a narrower claim than “calibration determines resale value,” and it holds up better. On TRS-RenTelco’s published 16-point certification checklist, only two points concern calibration at all — and one of those is confirming the tamper seal is intact. Panel condition, working controls, present fans and filters, verified options and complete accessories occupy the rest.
Where the buyer is a refurbisher, anyone preparing to sell test equipment out of a surplus lot usually does better describing the instrument accurately and letting the buyer price the calibration risk, rather than paying retail to remove it in advance.
Out of calibration is not the same as out of spec
A lapsed calibration date reads like a defect. It usually isn’t one.
The interval belongs to the owner, not the manufacturer. Tektronix’s own guidance on calibration intervals says there is “no ‘one size fits all’ answer” and that the interval “needs to be determined by the end user.” Fluke’s interval guidance goes further and lists “Never” as a legitimate tier for equipment used only for gross checks.
An instrument twenty months past its last calibration is out of somebody’s schedule. Whether it is out of tolerance is a separate question, and one nobody can answer without measuring.
What does signal a real problem is a failed signal path compensation — the routine a scope runs on itself to correct for temperature and time drift. Tektronix is blunt about what a failure there means: when an instrument reports an SPC failure, “99.9 percent of the time it is a hardware failure that will need repair.” Running it before listing anything separates instruments that will quote as working from instruments that will quote as repairs.
The value that isn’t visible from outside
Two other things move price more than most sellers realize, and neither shows up in a photograph.
Installed options. On current instruments, capability is licensed rather than built. A Tektronix 4 Series bandwidth upgrade from 200 MHz to 1.5 GHz lists at $27,600 and changes nothing physical. Serial decode and power analysis options run $2,890 to $3,940 each. An RSA5103B digital I/Q output option lists at $25,000.
Whether that value transfers depends on the license type, and the seller-advice guides skip it entirely. Tektronix node-locked licenses are “permanently enabled for the instrument” — they travel with the box. Floating licenses can be checked in and checked out to a different host. Keysight sells transportable and USB-portable licenses that behave the same way.
The practical consequence runs both directions. A transportable option can be pulled off before shipping, subject to whatever the specific license agreement says, so a buyer shouldn’t assume an advertised option string arrives with the box. A seller shouldn’t assume it either. Work out which options are welded to the instrument before listing, because only those belong in the asking price — and any that aren’t are worth retaining for the next instrument.
Probes and accessories. These carry more value than the instruments they attach to, more often than seems reasonable. A Tektronix TCP0030A current probe lists at $6,970 new; a THDP0200 high-voltage differential probe lists at $3,350. Those are single accessories priced above plenty of complete used mainframes.
Probes also outlive interfaces. Tektronix’s TPA-BNC adapter bridges TekProbe-era probes onto modern TekVPI inputs — an interface introduced in 1986 that still has a path onto current hardware. A probe drawer that looks like clutter is frequently the most liquid thing in the room.
The floor: when the manufacturer stops
Below a certain point an instrument stops being equipment and becomes parts, and the manufacturer sets that line rather than the hardware.
Tektronix publishes a four-stage support ladder: Full Support, then Calibration/Best Effort Repair, then Calibration Only, then End of Support. A unit at Calibration Only can still be certified but never repaired. At End of Support, neither.
Keysight draws a similar line with different words, and the words matter. A Discontinued product is no longer sold but remains fully supported, repair and calibration included. An Obsolete one gets “free self-support via the Keysight website only.”
Keysight’s lifecycle policy commits to support for up to five years after discontinuance. After that it continues case by case, “as long as parts and the required electronic test equipment are available.”
One qualifier keeps this from being a cliff. OEM end-of-support is not the same as nobody being able to calibrate the instrument. Independent labs advertise directly into that gap — Custom-Cal’s own marketing line is that obsolete instruments are not a problem, which is a sales claim rather than a market survey, but the business exists because the demand does. Losing factory service narrows the buyer pool. It rarely empties it.
The practical order, then, for anything sitting idle: run signal path compensation, find out where the model sits on the manufacturer’s support ladder, check which licenses are node-locked, and locate the probes. Those four answers determine the offer far more than the date on the calibration sticker.